Despite improvements in treatment modalities, approximately one-third of children with acute lymphoblastic leukemia (ALL), and two-thirds of patients with acute myeloid leukemia (AML) are not presently being cured. Following hematologic relapse, 80% of the patients expire within 6 months. At diagnosis, patients have a tumor burden of approximately one billion leukemic cells, and at the present time it is not known how effective cytoreductive therapy is. This is because residual leukemic blasts representing <5% of the bone marrow cell population may remain unseen by conventional morphological examination, and other methods of detection of residual leukemic cells are either insufficiently sensitive to significantly improve on this detection level, or require considerable technical expertise and can only be applied to small numbers of samples. It is possible that early detection of minimal residual disease (MRD) when the leukemic burden is low would allow intervention with alternative treatment strategies that may improve the patient's possibility of survival. The polymerase chain reaction (PCR) represents a technique that is potentially applicable to this problem. The sensitivity of the technique is sufficient to detect one cell in 100,000, and if specificity for the leukemic clone can be achieved then use of this technique will significantly improve our ability to detect occult leukemic cells. Leukemic clone specificity can be achieved through the use of immunoglobulin and T cell receptor gene rearrangements which can be found in almost 100% of B and T lineage ALLs, and in about 20% of AMLs. The dynamics of the leukemic cell population during remission will therefore be investigated utilizing the PCR to detect minimal residual disease. They will attempt to determine if detectable residual leukemic cells are a harbinger of impending relapse. Methods will be developed to make the PCR technique generally applicable to leukemias with gene rearrangements by identifying primers from immunoglobulin and T cell receptor genes that can be used for this purpose. With such primers, they will examine serial DNA samples from a large number of patients, some of whom remain in continuous remission and some of whom have relapsed. Patients undergoing autologous and allogeneic bone marrow transplantation will be examined before and after infusion of marrow. Overall, these studies should shed considerable light on the dynamics of the leukemic cell population during clinical remission, and determine whether the PCR can be used to predict impending relapse. A positive outcome of these experiments may allow future modifications in therapeutic modalities that are geared toward the actual level of leukemic cells in the body.